Comparative Efficiency of DBSCAN, K-means, and Their Combination in Image Processing

Charles Okanda Nyatega; Elizabeth Odrick Koola; Joseph Sospeter Salawa; Juma Said Ally; Cuthbert John Karawa; Phocas Sebastian; Richard Mwanjalila

doi:10.2022/ajest.v8i1.1076

Charles Okanda Nyatega
Elizabeth Odrick Koola Department of Electronics and Telecommunication Engineering, Mbeya University of Science and Technology (MUST)
Joseph Sospeter Salawa
Juma Said Ally
Cuthbert John Karawa
Phocas Sebastian
Richard Mwanjalila

DOI: https://doi.org/10.2022/ajest.v8i1.1076

##article.subject##: Clustering, DBSCAN, K-Means, Hybrid, MRI Images

##article.abstract##

Clustering techniques are vital for image analysis applications like object detection, pattern recognition, and image segmentation. The research evaluates effectiveness and efficiency of three clustering techniques K-means, DBSCAN (Density-Based Spatial Clustering of Applications with Noise), and a new hybrid model that combines the two are compared in this research in the context of analyzing MRI image dataset. Finding a technique that performs well across various kinds of image analysis applications in terms of computing efficiency and clustering quality is the main goal. In this study, a hybrid model is presented to overcome the shortcomings of individual techniques. The hybrid model aims at achieving a balanced performance that increases computational efficiency and clustering accuracy in MRI image analysis by combining the characteristics of DB- SCAN and K-means. This methodology makes use of DBSCAN’s capacity to detect clusters with complicated shapes and densities combined with K-means’ effectiveness in handling huge amounts of data with simple cluster structures, the research significance is on how it potentially changes how the information is extracted from MRI images using the hybrid technique also it advances the field of medical image analysis. Silhouette Score is an Evaluation metric employed to assess clustering quality, and execution time metric as Computational efficiency measure. Based to our research, the hybrid model achieves a balanced performance on analyzing variety images in dataset by utilizing the advantages of both K-means and DBSCAN with the aim of opening the door to more precise and effective analysis in the field of medical imaging as well as beyond.

References

Adangabe, A. A., Emmanuella, D. A., & Tigtig, J. (2021). Exploring the challenges facing teenage mothers in school and how they cope in a, a. (n.d.). Clustering in Machine learning - Image and video recognition. Re- trieved 2024-03-23, from https://www.linkedin.com/pulse/clustering-machine-learning-image-video-recognition-meghana-bandaru

Boulis, C., & Ostendorf, M. (2004). Combining Multiple Clustering Systems. In D. Hutchi- son et al. (Eds.), Knowledge Discovery in Databases: PKDD 2004 (Vol. 3202, pp. 63–74). Berlin, Heidelberg: Springer Berlin Heidelberg. Retrieved 2024-02-25, from http:// link.springer.com/10.1007/978-3-540-30116-5 9 (Series Title: Lecture Notes in Computer Science) doi: 10.1007/978-3-540-30116-5 9

Cluster Analysis. (2019, May). Retrieved 2024-06-23, from https://deepai.org/ machine-learning-glossary-and-terms/cluster-analysis
DBSCAN vs. K-Means: A Guide in Python. (n.d.). Retrieved 2024-02-27, from
https://www.newhorizons.com/resources/blog/dbscan-vs-kmeans-a-guide-in-python

Despotovic´, I., Goossens, B., & Philips, W. (2015). MRI Segmentation of the Human Brain: Chal- lenges, Methods, and Applications. Computational and Mathematical Methods in Medicine, 2015, 1–23. Retrieved 2024-02-26, from http://www.hindawi.com/journals/ cmmm/2015/450341/ doi: 10.1155/2015/450341

Ester, M., Kriegel, H., Sander, J., & Xu, X. (1996, August). A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise.. Retrieved 2024- 07-03, from https://www.semanticscholar.org/paper/A-Density-Based-Algorithm-for-Discovering-Clusters-Ester-Kriegel/ 5c8fe9a0412a078e30eb7e5eeb0068655b673e86

Guan, X., Zhao, Y., Nyatega, C. O., & Li, Q. (2023, April). Brain Tumor Segmentation Net-work with Multi-View Ensemble Discrimination and Kernel-Sharing Dilated Convolution. Brain Sciences, 13(4), 650. Retrieved 2024-07-11, from https://www.mdpi.com/ 2076-3425/13/4/650 (Number: 4 Publisher: Multidisciplinary Digital Publishing Institute) doi: 10.3390/brainsci13040650
Introduction to Image Segmentation with K-Means clustering. (n.d.). Retrieved 2024-07-03, fromhttps://www.kdnuggets.com/introduction-to-image-segmentation-with-k-means-clustering (Section: 2019 Aug Tutorials,
Overviews)

Jafarzadegan, M., Safi-Esfahani, F., & Beheshti, Z. (2019, December). Combining hierarchical clustering approaches using the PCA method. Expert Systems with Applications, 137, 1–10. Retrieved 2024-07-04, from https://www.sciencedirect.com/science/article/pii/S0957417419304737 doi: 10.1016/j.eswa.2019.06.064 Jain, R. (2012, May). A Hybrid Clustering Algorithm for Data Mining. Computer Science & Information Technology, 2. doi: 10.5121/csit.2012.2239

Khedairia, S., & Khadir, M. T. (2022, January). A multiple clustering combination ap- proach based on iterative voting process. Journal of King Saud University - Computer and Information Sciences, 34(1), 1370–1380. Retrieved 2024-02-25, from https://
linkinghub.elsevier.com/retrieve/pii/S131915781930597X doi: 10.1016/j.jksuci.2019.09.013 K-means Clustering Algorithm: Applications, Types, and Demos [Updated] Simplilearn. (n.d.). Retrieved 2024-02-27, from https://www.simplilearn.com/tutorials/ machine-learning-tutorial/k-means-clustering-algorithmMagnetic Resonance Imaging (MRI). (n.d.). Retrieved 2024-02-25, from https://
www.nibib.nih.gov/science-education/science-topics/magnetic-resonance-imaging-mri

Nyatega, C. O., Qiang, L., Adamu, M. J., & Kawuwa, H. B. (2022, October). Gray mat- ter, white matter and cerebrospinal fluid abnormalities in Parkinson’s disease: A voxel- based morphometry study. Frontiers in Psychiatry, 13, 1027907. Retrieved 2024-07- 11, from https://www.frontiersin.org/articles/10.3389/fpsyt.2022.1027907/full doi: 10.3389/fpsyt.2022.1027907

Nyatega, C. O., Qiang, L., Adamu, M. J., Younis, A., & Kawuwa, H. B. (2021, December). Altered Dynamic Functional Connectivity of Cuneus in Schizophrenia Patients: A Resting-State fMRI Study. Applied Sciences, 11(23), 11392. Retrieved 2024-07-11, from https:// www.mdpi.com/2076-3417/11/23/11392 doi: 10.3390/app112311392

Pedro, J. (2022, May). How to ensemble Clustering Algorithms. Retrieved 2024-02-28, from https://towardsdatascience.com/how-to-ensemble-clustering-algorithms-bf78d7602265

Sharma, E. (2020, June). K-Means vs. DBSCAN Clustering. Retrieved 2024-07-03, from
https://towardsdatascience.com/k-means-vs-dbscan-clustering-49f8e627de27

Strehl, A., & Ghosh, J. (n.d.). Cluster Ensembles – A Knowledge Reuse Framework for Combining Multiple Partitions.

Velunachiyar, S., & Sivakumar, K. (2023, June). Some Clustering Methods, Algorithms and their Applications. International Journal on Recent and Innovation Trends in Computing and Communication, 11(6s), 401–410. Retrieved 2024-02-26, from https://ijritcc
.org/index.php/ijritcc/article/view/6946 doi: 10.17762/ijritcc.v11i6s
.6946

Yan, J., & Liu, W. (2022, May). An Ensemble Clustering Approach (Consensus Cluster- ing) for High-Dimensional Data. Security and Communication Networks, 2022, 1–9. Retrieved 2024-02-26, from https://www.hindawi.com/journals/scn/2022/ 5629710/ doi: 10.1155/2022/5629710